Sökning: "Mats Brorsson"

Visar resultat 1 - 5 av 10 avhandlingar innehållade orden Mats Brorsson.

1. 1. Performance Characterization and Optimization of In-Memory Data Analytics on a Scale-up Server

   Författare :Ahsan Javed Awan; Eduard Ayguade; Mats Brorsson; Vladimir Vlassov; Lieven Eeckhout; KTH; []
   Nyckelord :TEKNIK OCH TEKNOLOGIER; ENGINEERING AND TECHNOLOGY; Workload Characterization; Big Data Analytics; Multicore Performance; Apache Spark; Near Data Processing; NUMA; Hyperthreading; Prefetchers; Coherently attached accelerators; Informations- och kommunikationsteknik; Information and Communication Technology;

   Sammanfattning : The sheer increase in the volume of data over the last decade has triggered research in cluster computing frameworks that enable web enterprises to extract big insights from big data. While Apache Spark defines the state of the art in big data analytics platforms for (i) exploiting data-flow and in-memory computing and (ii) for exhibiting superior scale-out performance on the commodity machines, little effort has been devoted to understanding the performance of in-memory data analytics with Spark on modern scale-up servers. LÄS MER

3. 2. Performance Characterization of In-Memory Data Analytics on a Scale-up Server

   Författare :Ahsan Javed Awan; Mats Brorsson; Vladimir Vlassov; Eduard Ayguade; Boris Grot; KTH; []
   Nyckelord :TEKNIK OCH TEKNOLOGIER; ENGINEERING AND TECHNOLOGY; Informations- och kommunikationsteknik; Information and Communication Technology;

   Sammanfattning : The sheer increase in volume of data over the last decade has triggered research in cluster computing frameworks that enable web enterprises to extract big insights from big data. While Apache Spark defines the state of the art in big data analytics platforms for (i) exploiting data-flow and in-memory computing and (ii) for exhibiting superior scale-out performance on the commodity machines, little effort has been devoted at understanding the performance of in-memory data analytics with Spark on modern scale-up servers. LÄS MER

4. 3. Techniques for Enhancing the Efficiency of Transactional Memory Systems

   Författare :Shady Issa; Paolo Romano; Vladimir Vlassov; Mats Brorsson; Konstantin Busch; KTH; []
   Nyckelord :TEKNIK OCH TEKNOLOGIER; ENGINEERING AND TECHNOLOGY; Transactional Memory; Parallel Programming; Concurrency Control; Self-tuning; Energy Efficiency; Data Partitioning; Dynamic Frequency and Voltage Scaling DVFS ; Hardware Transactional Memory HTM ; Memória Transacional; Programação Paralela; Eficiência Energética; Mecanismos de Auto-configuração; Partição de Dados; Frequência Dinâmica e do Escalonamento de Voltagem DVFS ; Memória Transacional em Hardware HTM ; Transaktionellt Minne; Parallellprogrammering; Samtida Exekvering; Självjusterande; Energieffektivitet; Datapartitionering; Dynamisk Frekvens och Volttalsskalning; Informations- och kommunikationsteknik; Information and Communication Technology;

   Sammanfattning : Transactional Memory (TM) is an emerging programming paradigm that drastically simplifies the development of concurrent applications by relieving programmers from a major source of complexity: how to ensure correct, yet efficient, synchronization of concurrent accesses to shared memory. Despite the large body of research devoted to this area, existing TM systems still suffer from severe limitations that hamper both their performance and energy efficiency. LÄS MER

5. 4. Parallel algorithms for real-time railway rescheduling

   Författare :Sai Prashanth Josyula; Johanna Törnquist Krasemann; Lars Lundberg; Mats Brorsson; Blekinge Tekniska Högskola; []
   Nyckelord :NATURVETENSKAP; NATURAL SCIENCES;

   Sammanfattning : In railway traffic systems, it is essential to achieve a high punctuality to satisfy the goals of the involved stakeholders. Thus, whenever disturbances occur, it is important to effectively reschedule trains while considering the perspectives of various stakeholders. LÄS MER

7. 5. Improving OpenMP Productivity with Data Locality Optimizations and High-resolution Performance Analysis

   Författare :Ananya Muddukrishna; Mats Brorsson; Nawal Copty; KTH; []
   Nyckelord :TEKNIK OCH TEKNOLOGIER; ENGINEERING AND TECHNOLOGY; OpenMP; Performance Analysis; Scheduling; Locality Optimizations;

   Sammanfattning : The combination of high-performance parallel programming and multi-core processors is the dominant approach to meet the ever increasing demand for computing performance today. The thesis is centered around OpenMP, a popular parallel programming API standard that enables programmers to quickly get started with writing parallel programs. LÄS MER